No, I shifted to much longer time-frames instead. Quick-hits in individual equities was the entire foundation of my career in the first place, but the juicy repeating patterns went away and never got replaced by new ones.
I understand the multitude of exchanges for stocks allows all sort of HFT gaming (modern forms of arbitraging) - not to mention exotic order-types that were created by some of these exchanges just to cater to HFT companies. I have a harder time understanding what type of gaming HFT companies can play for single-exchange products, like CL on CME Globex for example. Another thing I am curious about, specific to CME Globex ... for stop orders hosted by the CME exchange, how fast (or slow) is the trigger mechanism? If slow enough, that could allow collocated HFT market orders to get executed ahead of CM hosted stops, which IMO would be an unfair advantage. Thanks in advance
I can't answer this, nor can anyone really. My opinion would be that the old, predictable intra-day repeating patterns were inefficiencies that have been eliminated (hence they're 'old') by smarter (because they were 'predictable'), faster ('intra-day' and lower timeframes) traders (HFT's + others). One example that comes to mind is that I used to be able to detect laggards in the S&P based on intraday patterns and make trades on the assumption that they would follow the rest of the index/futures. That went away, as did many strategies over the years, with others that take their place. Most strategies that replace them were essentially capitalizing on inefficiencies in the market that are smaller (lowering profit margins), disappear faster (requiring faster technology), and sometimes require more complex analysis (smarter people) to detect, which makes sense from evolutional standpoint...survival of the fittest.
That's a good question. I don't know for certain, though I'm sure they'll tell you if you ring them up. CME in particular is very good at being transparent about their matching process. I would put money on this sequence of events: 1) Market trades at your trigger price. 2) Your order becomes a limit order and enters the queue at your limit price. 3) Other orders (HFT and others) enter the queue behind you. So the only way an HFT or anyone else can beat you to a price is if their order hits the matching engine before #2 happens (so before the price is traded at or at the same instant, before CME has been able to grab your order out of some type of storage to convert to a marketable limit order). I agree, if an HFT can jump in ahead of stop-limits already sitting at the matching engine that it would be unfair. However, there are limitations to any multi-threaded system (of which theirs is) and there could be a race condition causing other orders to enter ahead of yours at very small timescales. However, I would bet that it is impossible for another firm to react to that trade triggering your stop and THEN submit an order that beats out the trigger stop. That is when it would become truly unfair and unless CME is experiencing some major delays between sessions and at the matching engine this scenario would be highly highly unlikely.
Can you describe in general terms what methods you are using to make money utilizing HFT (ie latency arb, rebating, etc.) Also, what are some of the abusive practices (if any) that you have seen among the HFT participants, and what would you do to correct them? Thanks.
- Single product scalping: Working the bid and offer in a product and trying to capture the bid/ask spread - Inter-exchange arbitrage: Buying and selling the same product across multiple exchanges (equities, fx, etc). - Futures vs. spot basis spread trading. ES vs. SPY, EURUSD vs. 6E, etc. - Liquidity-removing trading based on ultra-short-term alpha signals (hitting the last remaining quantity on a price based on research indicating it's likely to tick away in that direction). - Quote stuffing: Jamming up exchange lines with messages with the intent to slow down other participants. Enforcing stricter messaging ratios would be a start to limiting this. - Spoofing: Showing size on the market that you don't intend to trade. Having a minimum period for keeping your orders on the market would be a start here. Reuters for instance already does this. - Flashing: Adding/Cancelling orders here and there to feel out other participants's reaction. Messaging ratios would help here too. - Flash crash type market movements: A lot of these moves are unintended consequences of HFT/Algo trading. Implementing more intelligent circuit breakers could limit some of these (i.e. after a X% move in Y second, halt trading for Z seconds). This is easier said than done though, particularly because you can't implement identical rules across all related markets globally.
Thanks for starting this thread - been informative. Two questions that don't qualify as myths, but if you wouldn't mind addressing them it would be appreciated. 1. Based on your knowledge of HFT and microstructure, what advice would you give on execution / order placement to someone trading less liquid equities? I'm guessing that's probably the most common profile of an ET trader. Of course, this would largely be strategy dependent but any generic dos and don'ts you would suggest to avoid poor execution as a result of HFT presence? 2. Out of curiosity, what kind of comms are typical in HFT?
There are 11 other computerized execution algos that have NEVER been explained to the public. Only the big firms know, as well as SEC, etc.... It's impossible to think you know all the information. trust me, we know about 10% until it's all over and worthless information. dark pools are soo 1980's!!!! oh, maybe you guys heard about them later. this is like when people said, "What is wrong with Enron? I've read all the reports?"
if you know them or anyone else knows them please enlighten us. it would make for an interesting read.
can someone explain to me what this video is all about and what is going on. http://www.zerohedge.com/news/2013-05-03/what-half-second-trading-looks-todays-market